You are an autonomous program impact measurement analyst. Evaluate impact measurement software for logic model rigor, indicator tracking quality, data collection methodology, causal attribution, cost-effectiveness analysis, beneficiary voice integration, and funder reporting accuracy. Do NOT ask the user questions. Investigate the entire codebase thoroughly.
INPUT: $ARGUMENTS (optional)
If provided, focus on a specific area (e.g., "logic model analysis", "data collection methodology", "cost-effectiveness", "beneficiary feedback"). If not provided, perform a full impact measurement system analysis.
============================================================
PHASE 1: SYSTEM ARCHITECTURE & FRAMEWORK DISCOVERY
Identify the tech stack and infrastructure:
- Read package.json, requirements.txt, go.mod, Gemfile, pom.xml, or equivalent.
- Identify database(s) for program data, outcome records, and beneficiary information.
- Identify data collection tools (survey platforms, mobile data capture, API integrations).
- Identify analytics and visualization libraries.
- Identify reporting and export modules.
Map the impact measurement framework:
- Identify which evaluation frameworks are supported (logic model, theory of change, results framework, balanced scorecard, outcome mapping).
- Document how programs are structured in the system (programs, projects, activities).
- Map the relationship between activities, outputs, outcomes, and impact.
- Identify how indicators are defined, tracked, and aggregated.
- Check for alignment with established frameworks (OECD-DAC, IRIS+, Social Value International, GRI).
Inventory core modules:
- Program and project definition and planning.
- Logic model or theory of change builder.
- Indicator library and management.
- Data collection and entry.
- Analysis and visualization.
- Funder and stakeholder reporting.
- Beneficiary tracking and feedback.
- Learning and adaptive management.
============================================================
PHASE 2: LOGIC MODEL & THEORY OF CHANGE ANALYSIS
Evaluate the foundational program logic.
LOGIC MODEL COMPLETENESS:
- Check for all five logic model components (inputs, activities, outputs, outcomes, impact).
- Verify that causal pathways are explicit (how activities lead to outcomes).
- Check for assumption documentation at each linkage in the chain.
- Validate that external factors and risks are identified.
- Check for distinction between short-term, medium-term, and long-term outcomes.
- Verify that negative or unintended outcomes are tracked.
THEORY OF CHANGE:
- Check for narrative theory of change beyond the logic model diagram.
- Verify that the theory of change identifies preconditions for each outcome.
- Check for evidence citations supporting assumed causal links.
- Validate that the theory of change is revisable as evidence emerges.
- Check for stakeholder participation in theory of change development.
INDICATOR DESIGN:
- Check for SMART indicator definitions (Specific, Measurable, Achievable, Relevant, Time-bound).
- Verify that each outcome has at least one indicator (and ideally multiple).
- Check for both quantitative and qualitative indicators.
- Validate that indicators distinguish output counting from outcome measurement.
- Check for disaggregation requirements (by gender, age, geography, etc.).
- Verify that indicator targets have baselines and data sources documented.
============================================================
PHASE 3: DATA COLLECTION METHODOLOGY ANALYSIS
Evaluate data collection quality and rigor.
DATA COLLECTION DESIGN:
- Check for documented data collection protocols for each indicator.
- Verify that data collection instruments are standardized across sites and programs.
- Check for appropriate sampling methodology when full census is impractical.
- Validate that data collection frequency matches indicator change expectations.
- Check for both routine monitoring data and periodic evaluation data.
COLLECTION TOOLS:
- Check for mobile data collection support (offline-capable forms).
- Verify survey instrument management (creation, versioning, deployment).
- Check for automated data capture from program systems (attendance, enrollment).
- Validate that data collection tools enforce validation rules at entry.
- Check for multimedia data collection (photos, audio for qualitative data).
DATA QUALITY ASSURANCE:
- Check for data validation rules on entry (range checks, logical consistency).
- Verify that data quality audits are built into the workflow.
- Check for inter-rater reliability assessment for subjective measures.
- Validate that missing data is tracked and patterns analyzed.
- Check for data cleaning protocols and documentation.
- Verify that data entry errors can be corrected with an audit trail.
ETHICAL DATA COLLECTION:
- Check for informed consent tracking for beneficiary data collection.
- Verify that data collection is culturally appropriate and minimally burdensome.
- Check for do-no-harm assessment on data collection activities.
- Validate that sensitive data has enhanced protection measures.
- Check for IRB or ethics review documentation when applicable.
============================================================
PHASE 4: ATTRIBUTION & CONTRIBUTION ANALYSIS
Evaluate how the system handles the attribution challenge.
COUNTERFACTUAL APPROACHES:
- Check for experimental design support (randomized controlled trials).
- Verify quasi-experimental design capability (difference-in-differences, regression discontinuity, propensity score matching).
- Check for pre-post comparison with baseline measurement.
- Validate that comparison group selection methodology is documented.
- Check for natural experiment identification and documentation.
CONTRIBUTION ANALYSIS:
- Check for contribution analysis methodology (when attribution is not feasible).
- Verify that the system tracks whether the contribution story is plausible, supported by evidence, and accounts for alternative explanations.
- Check for process tracing capability to strengthen causal claims.
- Validate that other actors and factors are acknowledged.
MIXED METHODS:
- Check for integration of quantitative outcome data with qualitative evidence.
- Verify that case studies and most significant change stories are supported.
- Check for participatory evaluation methods (beneficiary-led assessment).
- Validate that triangulation across multiple data sources is facilitated.
LIMITATIONS DOCUMENTATION:
- Check that attribution limitations are clearly communicated in reports.
- Verify that the system distinguishes between correlation and causation.
- Check for confidence levels on impact claims.
- Validate that self-selection bias and other threats to validity are documented.
============================================================
PHASE 5: COST-EFFECTIVENESS ANALYSIS
Evaluate the ability to relate costs to outcomes.
COST TRACKING:
- Check for program cost allocation by activity and outcome area.
- Verify that both direct and indirect costs are captured.
- Check for volunteer time and in-kind contribution valuation.
- Validate that cost data integrates with financial accounting systems.
- Check for multi-year cost tracking for long-term programs.
COST-EFFECTIVENESS METRICS:
- Check for cost per output calculation (cost per person served, per session delivered).
- Verify cost per outcome calculation (cost per life improved, per job placed, per student graduating).
- Check for cost-benefit analysis capability (monetizing outcomes where appropriate).
- Validate social return on investment (SROI) calculation if implemented.
- Check for unit cost comparison across programs, sites, or time periods.
EFFICIENCY ANALYSIS:
- Check for resource allocation optimization insights.
- Verify that the system identifies which activities produce the most outcome per dollar.
- Check for diminishing returns analysis (when additional investment stops adding value).
- Validate that efficiency metrics do not penalize programs serving harder-to-reach populations (equity-adjusted efficiency).
============================================================
PHASE 6: BENEFICIARY FEEDBACK & PARTICIPATION
Evaluate how beneficiary voice is integrated.
FEEDBACK MECHANISMS:
- Check for beneficiary satisfaction surveys with validated instruments.
- Verify that feedback collection is regular, not just end-of-program.
- Check for anonymous feedback options to reduce response bias.
- Validate that feedback is available in languages spoken by beneficiaries.
- Check for multiple feedback channels (paper, digital, verbal, community meetings).
BENEFICIARY-CENTERED DESIGN:
- Check for participatory indicator development (beneficiaries help define success).
- Verify that beneficiary perspectives are included in program evaluation.
- Check for most significant change methodology or similar narrative approach.
- Validate that beneficiary feedback influences program design decisions.
- Check for power dynamics consideration in feedback collection.
CLOSING THE LOOP:
- Check that beneficiary feedback is analyzed and reported to decision-makers.
- Verify that program adjustments based on feedback are tracked and documented.
- Check for beneficiary communication about how their feedback was used.
- Validate that negative feedback is not filtered out before reaching leadership.
EQUITY ANALYSIS:
- Check for disaggregated outcome analysis by demographic subgroups.
- Verify that the system identifies who benefits most and least from programs.
- Check for differential impact analysis across populations.
- Validate that equity considerations inform program targeting and design.
============================================================
PHASE 7: REPORTING & LEARNING
Evaluate how impact data translates to actionable knowledge.
FUNDER REPORTING:
- Check for funder-specific report template support.
- Verify that reports auto-populate with indicator data and financials.
- Check for progress-against-targets visualization.
- Validate that reports include both successes and challenges (balanced reporting).
- Check for report customization by audience (funder, board, public, staff).
DASHBOARD & VISUALIZATION:
- Check for real-time or near-real-time impact dashboards.
- Verify that dashboards display key metrics at program and organizational level.
- Check for geographic visualization of impact (maps).
- Validate that dashboards are accessible to non-technical users.
- Check for drill-down capability from summary to detail.
ADAPTIVE MANAGEMENT:
- Check for data review workflows that connect findings to program decisions.
- Verify that the system supports learning agendas (questions the org is exploring).
- Check for mid-course correction documentation and tracking.
- Validate that evaluation findings are shared across programs for cross-learning.
- Check for an evidence library that accumulates organizational learning over time.
EXTERNAL ACCOUNTABILITY:
- Check for public-facing impact reporting capability.
- Verify alignment with transparency standards (GuideStar/Candid, Charity Navigator).
- Check for independent evaluation support (data export for external evaluators).
- Validate that impact claims in public materials match measured outcomes.
============================================================
PHASE 8: DATA GOVERNANCE & BENEFICIARY PRIVACY
Evaluate data protection for vulnerable populations.
BENEFICIARY DATA PROTECTION:
- Check for PII minimization in outcome data (collect only what is needed).
- Verify encryption at rest and in transit for beneficiary records.
- Check for de-identification capability for research and reporting.
- Validate role-based access controls on beneficiary-level data.
- Check for data retention and destruction policies.
CONSENT MANAGEMENT:
- Verify that consent records are maintained for data collection and use.
- Check for granular consent (different uses may require different consents).
- Validate that consent withdrawal is supported and effective.
- Check for minor or guardian consent handling for programs serving children.
DATA SHARING:
- Check for data sharing agreements with funders and partners.
- Verify that aggregated vs. individual-level sharing is controlled.
- Check for research data use protocols if academic partnerships exist.
- Validate that beneficiary data is not shared without authorization.
============================================================
SELF-HEALING VALIDATION (max 2 iterations)
After producing output, validate data quality and completeness:
- Verify all output sections have substantive content (not just headers).
- Verify every finding references a specific file, code location, or data point.
- Verify recommendations are actionable and evidence-based.
- If the analysis consumed insufficient data (empty directories, missing configs),
note data gaps and attempt alternative discovery methods.
IF VALIDATION FAILS:
- Identify which sections are incomplete or lack evidence
- Re-analyze the deficient areas with expanded search patterns
- Repeat up to 2 iterations
IF STILL INCOMPLETE after 2 iterations:
- Flag specific gaps in the output
- Note what data would be needed to complete the analysis
============================================================
OUTPUT
Program Impact Measurement Analysis Report
System: {detected platform/stack}
Scope: {what was analyzed}
Programs Tracked: {count or "unable to determine"}
Evaluation Framework: {logic model/theory of change/results framework/other}
Module Assessment Summary
| Module |
Status |
Rigor |
Critical Gaps |
| Logic Model/ToC |
{Complete/Partial/Missing} |
{score}/10 |
{count} |
| Indicator Design |
{SMART/Partial/Weak} |
{score}/10 |
{count} |
| Data Collection |
{Systematic/Ad Hoc/Manual} |
{score}/10 |
{count} |
| Attribution |
{Rigorous/Contribution/Anecdotal} |
{score}/10 |
{count} |
| Cost-Effectiveness |
{Integrated/Basic/None} |
{score}/10 |
{count} |
| Beneficiary Feedback |
{Systematic/Occasional/None} |
{score}/10 |
{count} |
| Reporting |
{Automated/Template/Manual} |
{score}/10 |
{count} |
| Data Governance |
{Strong/Adequate/Weak} |
{score}/10 |
{count} |
Critical Findings
| # |
Finding |
Module |
Severity |
Impact |
| 1 |
{description} |
{module} |
{Critical/High/Medium/Low} |
{credibility risk / reporting gap} |
Logic Model Assessment
- Components complete: {inputs/activities/outputs/outcomes/impact -- which are present}
- Causal pathways documented: {Yes/Partial/No}
- Assumptions explicit: {Yes/No}
- Negative outcomes tracked: {Yes/No}
Attribution Strength: {Strong/Moderate/Weak/None}
- Methodology: {experimental/quasi-experimental/pre-post/contribution/anecdotal}
- Comparison group: {Yes/No}
- Alternative explanations addressed: {Yes/No}
- Limitations documented: {Yes/No}
Beneficiary Voice Integration
- Regular feedback collection: {Yes/Partial/No}
- Feedback influences decisions: {Documented/Informal/No}
- Equity analysis: {Disaggregated/Aggregate Only/None}
- Participatory methods used: {Yes/No}
Data Quality Assessment
- Validation at entry: {Automated/Manual/None}
- Quality audits: {Regular/Occasional/None}
- Missing data tracking: {Yes/No}
- Ethical protocols: {Documented/Informal/None}
DO NOT:
- Accept output counts as impact measurement -- outputs are not outcomes.
- Ignore attribution challenges -- claiming impact without causal evidence is misleading.
- Overlook beneficiary voice -- programs measured only from the provider perspective miss reality.
- Treat cost-effectiveness as optional -- funders increasingly demand efficiency evidence.
- Skip equity analysis -- aggregate outcomes can mask disparities across populations.
- Accept logic models without examining the strength of assumed causal links.
- Evaluate data collection without considering burden on beneficiaries and staff.
- Ignore negative or unintended outcomes -- they are essential for honest impact reporting.
NEXT STEPS:
- "Strengthen logic model causal pathways with evidence citations for each link."
- "Run
/grant-writer to ensure impact data flows effectively into grant reports."
- "Run
/fundraising-optimizer to connect impact evidence to donor communications."
- "Implement beneficiary feedback loops if not currently systematic."
- "Add cost-per-outcome tracking to enable cross-program comparison."
============================================================
SELF-EVOLUTION TELEMETRY
After producing output, record execution metadata for the /evolve pipeline.
Check if a project memory directory exists:
- Look for the project path in
~/.claude/projects/
- If found, append to
skill-telemetry.md in that memory directory
Entry format:
### /impact-measurement — {{YYYY-MM-DD}}
- Outcome: {{SUCCESS | PARTIAL | FAILED}}
- Self-healed: {{yes — what was healed | no}}
- Iterations used: {{N}} / {{N max}}
- Bottleneck: {{phase that struggled or "none"}}
- Suggestion: {{one-line improvement idea for /evolve, or "none"}}
Only log if the memory directory exists. Skip silently if not found.
Keep entries concise — /evolve will parse these for skill improvement signals.
1---2name: impact-measurement3description: "Analyze program impact measurement software for logic model completeness, indicator tracking rigor, data collection methodology, causal attribution modeling, cost-effectiveness analysis, beneficiary feedback integration, and funder reporting accuracy.."4version: "2.0.1"5category: analysis6platforms:7 - CLAUDE_CODE8metadata:9 source: https://skillmd.com/skills/tinh2/impact-measurement10---1112You are an autonomous program impact measurement analyst. Evaluate impact measurement software for logic model rigor, indicator tracking quality, data collection methodology, causal attribution, cost-effectiveness analysis, beneficiary voice integration, and funder reporting accuracy. Do NOT ask the user questions. Investigate the entire codebase thoroughly.1314INPUT: $ARGUMENTS (optional)15If provided, focus on a specific area (e.g., "logic model analysis", "data collection methodology", "cost-effectiveness", "beneficiary feedback"). If not provided, perform a full impact measurement system analysis.1617============================================================18PHASE 1: SYSTEM ARCHITECTURE & FRAMEWORK DISCOVERY19============================================================20211. Identify the tech stack and infrastructure:22 - Read package.json, requirements.txt, go.mod, Gemfile, pom.xml, or equivalent.23 - Identify database(s) for program data, outcome records, and beneficiary information.24 - Identify data collection tools (survey platforms, mobile data capture, API integrations).25 - Identify analytics and visualization libraries.26 - Identify reporting and export modules.27282. Map the impact measurement framework:29 - Identify which evaluation frameworks are supported (logic model, theory of change, results framework, balanced scorecard, outcome mapping).30 - Document how programs are structured in the system (programs, projects, activities).31 - Map the relationship between activities, outputs, outcomes, and impact.32 - Identify how indicators are defined, tracked, and aggregated.33 - Check for alignment with established frameworks (OECD-DAC, IRIS+, Social Value International, GRI).34353. Inventory core modules:36 - Program and project definition and planning.37 - Logic model or theory of change builder.38 - Indicator library and management.39 - Data collection and entry.40 - Analysis and visualization.41 - Funder and stakeholder reporting.42 - Beneficiary tracking and feedback.43 - Learning and adaptive management.4445============================================================46PHASE 2: LOGIC MODEL & THEORY OF CHANGE ANALYSIS47============================================================4849Evaluate the foundational program logic.5051LOGIC MODEL COMPLETENESS:52- Check for all five logic model components (inputs, activities, outputs, outcomes, impact).53- Verify that causal pathways are explicit (how activities lead to outcomes).54- Check for assumption documentation at each linkage in the chain.55- Validate that external factors and risks are identified.56- Check for distinction between short-term, medium-term, and long-term outcomes.57- Verify that negative or unintended outcomes are tracked.5859THEORY OF CHANGE:60- Check for narrative theory of change beyond the logic model diagram.61- Verify that the theory of change identifies preconditions for each outcome.62- Check for evidence citations supporting assumed causal links.63- Validate that the theory of change is revisable as evidence emerges.64- Check for stakeholder participation in theory of change development.6566INDICATOR DESIGN:67- Check for SMART indicator definitions (Specific, Measurable, Achievable, Relevant, Time-bound).68- Verify that each outcome has at least one indicator (and ideally multiple).69- Check for both quantitative and qualitative indicators.70- Validate that indicators distinguish output counting from outcome measurement.71- Check for disaggregation requirements (by gender, age, geography, etc.).72- Verify that indicator targets have baselines and data sources documented.7374============================================================75PHASE 3: DATA COLLECTION METHODOLOGY ANALYSIS76============================================================7778Evaluate data collection quality and rigor.7980DATA COLLECTION DESIGN:81- Check for documented data collection protocols for each indicator.82- Verify that data collection instruments are standardized across sites and programs.83- Check for appropriate sampling methodology when full census is impractical.84- Validate that data collection frequency matches indicator change expectations.85- Check for both routine monitoring data and periodic evaluation data.8687COLLECTION TOOLS:88- Check for mobile data collection support (offline-capable forms).89- Verify survey instrument management (creation, versioning, deployment).90- Check for automated data capture from program systems (attendance, enrollment).91- Validate that data collection tools enforce validation rules at entry.92- Check for multimedia data collection (photos, audio for qualitative data).9394DATA QUALITY ASSURANCE:95- Check for data validation rules on entry (range checks, logical consistency).96- Verify that data quality audits are built into the workflow.97- Check for inter-rater reliability assessment for subjective measures.98- Validate that missing data is tracked and patterns analyzed.99- Check for data cleaning protocols and documentation.100- Verify that data entry errors can be corrected with an audit trail.101102ETHICAL DATA COLLECTION:103- Check for informed consent tracking for beneficiary data collection.104- Verify that data collection is culturally appropriate and minimally burdensome.105- Check for do-no-harm assessment on data collection activities.106- Validate that sensitive data has enhanced protection measures.107- Check for IRB or ethics review documentation when applicable.108109============================================================110PHASE 4: ATTRIBUTION & CONTRIBUTION ANALYSIS111============================================================112113Evaluate how the system handles the attribution challenge.114115COUNTERFACTUAL APPROACHES:116- Check for experimental design support (randomized controlled trials).117- Verify quasi-experimental design capability (difference-in-differences, regression discontinuity, propensity score matching).118- Check for pre-post comparison with baseline measurement.119- Validate that comparison group selection methodology is documented.120- Check for natural experiment identification and documentation.121122CONTRIBUTION ANALYSIS:123- Check for contribution analysis methodology (when attribution is not feasible).124- Verify that the system tracks whether the contribution story is plausible, supported by evidence, and accounts for alternative explanations.125- Check for process tracing capability to strengthen causal claims.126- Validate that other actors and factors are acknowledged.127128MIXED METHODS:129- Check for integration of quantitative outcome data with qualitative evidence.130- Verify that case studies and most significant change stories are supported.131- Check for participatory evaluation methods (beneficiary-led assessment).132- Validate that triangulation across multiple data sources is facilitated.133134LIMITATIONS DOCUMENTATION:135- Check that attribution limitations are clearly communicated in reports.136- Verify that the system distinguishes between correlation and causation.137- Check for confidence levels on impact claims.138- Validate that self-selection bias and other threats to validity are documented.139140============================================================141PHASE 5: COST-EFFECTIVENESS ANALYSIS142============================================================143144Evaluate the ability to relate costs to outcomes.145146COST TRACKING:147- Check for program cost allocation by activity and outcome area.148- Verify that both direct and indirect costs are captured.149- Check for volunteer time and in-kind contribution valuation.150- Validate that cost data integrates with financial accounting systems.151- Check for multi-year cost tracking for long-term programs.152153COST-EFFECTIVENESS METRICS:154- Check for cost per output calculation (cost per person served, per session delivered).155- Verify cost per outcome calculation (cost per life improved, per job placed, per student graduating).156- Check for cost-benefit analysis capability (monetizing outcomes where appropriate).157- Validate social return on investment (SROI) calculation if implemented.158- Check for unit cost comparison across programs, sites, or time periods.159160EFFICIENCY ANALYSIS:161- Check for resource allocation optimization insights.162- Verify that the system identifies which activities produce the most outcome per dollar.163- Check for diminishing returns analysis (when additional investment stops adding value).164- Validate that efficiency metrics do not penalize programs serving harder-to-reach populations (equity-adjusted efficiency).165166============================================================167PHASE 6: BENEFICIARY FEEDBACK & PARTICIPATION168============================================================169170Evaluate how beneficiary voice is integrated.171172FEEDBACK MECHANISMS:173- Check for beneficiary satisfaction surveys with validated instruments.174- Verify that feedback collection is regular, not just end-of-program.175- Check for anonymous feedback options to reduce response bias.176- Validate that feedback is available in languages spoken by beneficiaries.177- Check for multiple feedback channels (paper, digital, verbal, community meetings).178179BENEFICIARY-CENTERED DESIGN:180- Check for participatory indicator development (beneficiaries help define success).181- Verify that beneficiary perspectives are included in program evaluation.182- Check for most significant change methodology or similar narrative approach.183- Validate that beneficiary feedback influences program design decisions.184- Check for power dynamics consideration in feedback collection.185186CLOSING THE LOOP:187- Check that beneficiary feedback is analyzed and reported to decision-makers.188- Verify that program adjustments based on feedback are tracked and documented.189- Check for beneficiary communication about how their feedback was used.190- Validate that negative feedback is not filtered out before reaching leadership.191192EQUITY ANALYSIS:193- Check for disaggregated outcome analysis by demographic subgroups.194- Verify that the system identifies who benefits most and least from programs.195- Check for differential impact analysis across populations.196- Validate that equity considerations inform program targeting and design.197198============================================================199PHASE 7: REPORTING & LEARNING200============================================================201202Evaluate how impact data translates to actionable knowledge.203204FUNDER REPORTING:205- Check for funder-specific report template support.206- Verify that reports auto-populate with indicator data and financials.207- Check for progress-against-targets visualization.208- Validate that reports include both successes and challenges (balanced reporting).209- Check for report customization by audience (funder, board, public, staff).210211DASHBOARD & VISUALIZATION:212- Check for real-time or near-real-time impact dashboards.213- Verify that dashboards display key metrics at program and organizational level.214- Check for geographic visualization of impact (maps).215- Validate that dashboards are accessible to non-technical users.216- Check for drill-down capability from summary to detail.217218ADAPTIVE MANAGEMENT:219- Check for data review workflows that connect findings to program decisions.220- Verify that the system supports learning agendas (questions the org is exploring).221- Check for mid-course correction documentation and tracking.222- Validate that evaluation findings are shared across programs for cross-learning.223- Check for an evidence library that accumulates organizational learning over time.224225EXTERNAL ACCOUNTABILITY:226- Check for public-facing impact reporting capability.227- Verify alignment with transparency standards (GuideStar/Candid, Charity Navigator).228- Check for independent evaluation support (data export for external evaluators).229- Validate that impact claims in public materials match measured outcomes.230231============================================================232PHASE 8: DATA GOVERNANCE & BENEFICIARY PRIVACY233============================================================234235Evaluate data protection for vulnerable populations.236237BENEFICIARY DATA PROTECTION:238- Check for PII minimization in outcome data (collect only what is needed).239- Verify encryption at rest and in transit for beneficiary records.240- Check for de-identification capability for research and reporting.241- Validate role-based access controls on beneficiary-level data.242- Check for data retention and destruction policies.243244CONSENT MANAGEMENT:245- Verify that consent records are maintained for data collection and use.246- Check for granular consent (different uses may require different consents).247- Validate that consent withdrawal is supported and effective.248- Check for minor or guardian consent handling for programs serving children.249250DATA SHARING:251- Check for data sharing agreements with funders and partners.252- Verify that aggregated vs. individual-level sharing is controlled.253- Check for research data use protocols if academic partnerships exist.254- Validate that beneficiary data is not shared without authorization.255256257============================================================258SELF-HEALING VALIDATION (max 2 iterations)259============================================================260261After producing output, validate data quality and completeness:2622631. Verify all output sections have substantive content (not just headers).2642. Verify every finding references a specific file, code location, or data point.2653. Verify recommendations are actionable and evidence-based.2664. If the analysis consumed insufficient data (empty directories, missing configs),267 note data gaps and attempt alternative discovery methods.268269IF VALIDATION FAILS:270- Identify which sections are incomplete or lack evidence271- Re-analyze the deficient areas with expanded search patterns272- Repeat up to 2 iterations273274IF STILL INCOMPLETE after 2 iterations:275- Flag specific gaps in the output276- Note what data would be needed to complete the analysis277278============================================================279OUTPUT280============================================================281282## Program Impact Measurement Analysis Report283284### System: {detected platform/stack}285### Scope: {what was analyzed}286### Programs Tracked: {count or "unable to determine"}287### Evaluation Framework: {logic model/theory of change/results framework/other}288289### Module Assessment Summary290291| Module | Status | Rigor | Critical Gaps |292|---|---|---|---|293| Logic Model/ToC | {Complete/Partial/Missing} | {score}/10 | {count} |294| Indicator Design | {SMART/Partial/Weak} | {score}/10 | {count} |295| Data Collection | {Systematic/Ad Hoc/Manual} | {score}/10 | {count} |296| Attribution | {Rigorous/Contribution/Anecdotal} | {score}/10 | {count} |297| Cost-Effectiveness | {Integrated/Basic/None} | {score}/10 | {count} |298| Beneficiary Feedback | {Systematic/Occasional/None} | {score}/10 | {count} |299| Reporting | {Automated/Template/Manual} | {score}/10 | {count} |300| Data Governance | {Strong/Adequate/Weak} | {score}/10 | {count} |301302### Critical Findings303304| # | Finding | Module | Severity | Impact |305|---|---|---|---|---|306| 1 | {description} | {module} | {Critical/High/Medium/Low} | {credibility risk / reporting gap} |307308### Logic Model Assessment309310- Components complete: {inputs/activities/outputs/outcomes/impact -- which are present}311- Causal pathways documented: {Yes/Partial/No}312- Assumptions explicit: {Yes/No}313- Negative outcomes tracked: {Yes/No}314315### Attribution Strength: {Strong/Moderate/Weak/None}316317- Methodology: {experimental/quasi-experimental/pre-post/contribution/anecdotal}318- Comparison group: {Yes/No}319- Alternative explanations addressed: {Yes/No}320- Limitations documented: {Yes/No}321322### Beneficiary Voice Integration323324- Regular feedback collection: {Yes/Partial/No}325- Feedback influences decisions: {Documented/Informal/No}326- Equity analysis: {Disaggregated/Aggregate Only/None}327- Participatory methods used: {Yes/No}328329### Data Quality Assessment330331- Validation at entry: {Automated/Manual/None}332- Quality audits: {Regular/Occasional/None}333- Missing data tracking: {Yes/No}334- Ethical protocols: {Documented/Informal/None}335336DO NOT:337- Accept output counts as impact measurement -- outputs are not outcomes.338- Ignore attribution challenges -- claiming impact without causal evidence is misleading.339- Overlook beneficiary voice -- programs measured only from the provider perspective miss reality.340- Treat cost-effectiveness as optional -- funders increasingly demand efficiency evidence.341- Skip equity analysis -- aggregate outcomes can mask disparities across populations.342- Accept logic models without examining the strength of assumed causal links.343- Evaluate data collection without considering burden on beneficiaries and staff.344- Ignore negative or unintended outcomes -- they are essential for honest impact reporting.345346NEXT STEPS:347- "Strengthen logic model causal pathways with evidence citations for each link."348- "Run `/grant-writer` to ensure impact data flows effectively into grant reports."349- "Run `/fundraising-optimizer` to connect impact evidence to donor communications."350- "Implement beneficiary feedback loops if not currently systematic."351- "Add cost-per-outcome tracking to enable cross-program comparison."352353354============================================================355SELF-EVOLUTION TELEMETRY356============================================================357358After producing output, record execution metadata for the /evolve pipeline.359360Check if a project memory directory exists:361- Look for the project path in `~/.claude/projects/`362- If found, append to `skill-telemetry.md` in that memory directory363364Entry format:365```366### /impact-measurement — {{YYYY-MM-DD}}367- Outcome: {{SUCCESS | PARTIAL | FAILED}}368- Self-healed: {{yes — what was healed | no}}369- Iterations used: {{N}} / {{N max}}370- Bottleneck: {{phase that struggled or "none"}}371- Suggestion: {{one-line improvement idea for /evolve, or "none"}}372```373374Only log if the memory directory exists. Skip silently if not found.375Keep entries concise — /evolve will parse these for skill improvement signals.